a new lhc device for the lhc: implementation aspects/options/cost estimates

Beam-gas Imaging for LHC – kickoff meeting 30-Oct-2012 CERN Massimiliano Ferro-Luzzi 1

A new LHC device for the LHC: implementation aspects/options/cost

estimates

Detector Readout chain Gas target organization


the detectorthe detector


New Device for LHCNew Device for LHC

Specifications:Specifications:provide measurement of emittance

– stat accuracy of <5% in 5 min for 1011 p

– syst uncertainty <5% (dominated by understanding of vtx resolution ?)

– for 1.5 < n/um < 4 and 0.45 < E/TeV < 7

– bunch by bunch (max bunches per measurement to be defined?)should not affect beam operation

By-products:By-products: measurement of ghost charge (crucial for lumi calib) see A. Alici et al. (BCNWG note4), CERN-ATS-Note-2012-029 PERF

measurement of relative bunch charges (to be normed by DCCT) see G. Anders et al. (BCNWG note3), CERN-ATS-Note-2012-028 PERF

Possible add-on: Possible add-on: (not in baseline discussed here)timing detector with < ~100ps resolution, would provide longitudinal profile as well

NB: it will measure beam shapes → must be able to measure accurately absolute value of function at the device position!

NB: it will measure beam shapes → must be able to measure accurately absolute value of function at the device position!


Starting point (educated guess)Starting point (educated guess)

Z range to be covered by acceptance Take: L = 1mZ range to be covered by acceptance Take: L = 1m

Accept: 3 < < 5 ( 100 > /mrad > 15 )Accept: 3 < < 5 ( 100 > /mrad > 15 )

R1min

max

R2

z1 = L/2 + R1 / tan(min) Take: R1 = 15, 20, 25 mm Rw = (z1 + L/2) tan(max)R2 = (z2 + L/2) tan(max)

z1 = L/2 + R1 / tan(min) Take: R1 = 15, 20, 25 mm Rw = (z1 + L/2) tan(max)R2 = (z2 + L/2) tan(max)

( VELO: R1 = 8mm, L = ~0.5-1m )

∆z-L/2

First detector Last detector

z1 z20

windowradius

Rw

given by beam pipe

• small R1 implies small (compact) detector (probably silicon)• larger R1 implies larger detector (probably SciFi)

• small R1 implies small (compact) detector (probably silicon)• larger R1 implies larger detector (probably SciFi)

L/2


JUST AN EXAMPLEJUST AN EXAMPLE

Simplest solution with straight segmented detectorsSimplest solution with straight segmented detectors

H

all tracks within radius R2 are acceptedR2

2 = R12 + (H-R1)2

+ some bonus outside, in the corners...

2R1

R2

beam pipe

x y x’ y’

z

«minimum» number of planes (4)

x’’ y’’ x y


Comments to geometryComments to geometry

We must find the place with the smallest ratio RWe must find the place with the smallest ratio Rpipepipe//beambeam

Want small aperture, but large beams!! :-)Option not considered here: movable device (in vacuum)

– complexity (and cost) increases substantially, but allows to come closer

What is the smallest possible radius ? What is the smallest possible radius ? (even if smaller beam size)Drives whether we should use silicon or SciFi

Performance/cost optimization to be donePerformance/cost optimization to be doneInclined detectors, stereo angle, ...Saw tooth beam pipe ?Put ring1 device such that no bkg to ring2 device (and vice versa) Can we live with 4 planes (xyx’y’) using vertex-constrained tracking ?etc...


recycle from LHCb ?recycle from LHCb ?

silicon trackers (straight strips)silicon trackers (straight strips)

TTTT ITIT

BOTH TO BE REPLACED AT LS2(LHCb Upgrade)

BOTH TO BE REPLACED AT LS2(LHCb Upgrade)


LHCb TT in more detailsLHCb TT in more details

34 staves/plane x 4 planes = 136 staves Cooled to about 0 oC Issue: sensors are glued on the stave (7 by 7)

One plane (out of four)One plane (out of four) One stave (half module)One stave (half module)

about 70cm

about 70cm

about

10cmabout

10cm

Sensor thickness pitch dim (mm2)500 um 183um 96.4x94.4

Sensor thickness pitch dim (mm2)500 um 183um 96.4x94.4


LHCb IT in more detailsLHCb IT in more details

14 osm/plane x 12 planes = 168 osm 14 tsm/plane x 12 planes = 168 tsm Cooled to about 0 oC

One plane (out of twelve)One plane (out of twelve) One moduleOne module

module type thickness pitch dim (mm2)one-sensor modules (osm) 320 um 198um 76x110two-sensor modules (tsm) 410 um 198 um 76x110

module type thickness pitch dim (mm2)one-sensor modules (osm) 320 um 198um 76x110two-sensor modules (tsm) 410 um 198 um 76x110


Silicon modules in LHCbSilicon modules in LHCb

TT (Tracker Turicensis)(or Trigger Tracker)TT (Tracker Turicensis)(or Trigger Tracker) IT (Inner Tracker)IT (Inner Tracker)

1,2,3- or 4-sensor high1,2,3- or 4-sensor high 1- or 2-sensor high1- or 2-sensor high

Both working well Apart perhaps from central part of

TT, the detector modules will have little radiation damage, but will be activated (lightly radioactive material)


Scintillating Fiber modules in LHCb ?Scintillating Fiber modules in LHCb ?

No SciFi detector in LHCb currently, but... Important R&D effort ongoing to consider SciFi for upgrade

– would replace central part of Outer Tracker and the Inner Tracker Already some short modules available as a spare for the IT boxes

– ~11cm long fibers More modules being fabricated for Upgrade R&D, with several

lengths (22cm, 80cm, 250cm) Groups: EPFL, Dortmund, Heidelberg, CERN, ...


Movable vs not movableMovable vs not movable

Movable detector device:Movable detector device:can come closer to beam

– improves resolution, compactifies the detectorincreases complexity of detector

– mechanics, vacuum, protection, etc...Issues:

– only movable in StableBeams ?

– difficult to access detector (repair, maintenance)

My personal bias: «if it can be done with a fixed detector, don’t make it movable…»


Possible add-on to be considered/studiedPossible add-on to be considered/studied

Currently, in LHCb: 25ns integration time, 40MHz synchronous– OK for nominal RF buckets

– But for ghost charge, had to check detection efficiency vs time

Fast timing detectors Fast timing detectors

See e.g. AFP: based on cerenkov radiation, quartz rods with MCP-PMT can reach ~ 20ps

Would allow measuring the longitudinal profile– satellites, bunch lengths, ...

– Scintillators with ~ 0.3 ns are sufficient for satellite measurements, but we need < ~100 ps for proper bunch length measurement

25ns slot

Nominal rf bucket ghost charge anywhere

See Plamen’s thesisSee Plamen’s thesis


readout chainreadout chain


IT/TT FE electronicsIT/TT FE electronics

In principle, all these beautiful electronics are to be «scrapped» at LS2...


LHCb readout chain (IT and TT)LHCb readout chain (IT and TT)

Now

Upgrade

40MHz data ZS40MHz data ZS

max 1MHz data ZSmax 1MHz data ZSnetwork

CPU farm

FE electronics Readout board

L0 electronics

near detector ... 100m ... electronics racknear detector ... 100m ... electronics rack

1-40MHz1-40MHz

1 MHz1 MHz


At the LHC, a simple view...At the LHC, a simple view...

FE electronics

Hw trigger decision

trackingdetectors

timing/trigdetectors

readoutboards

gas target

readoutnetwork

CPU farm

tunnel

LHCb upgrade


Data rate (to be refined at a later stage…)Data rate (to be refined at a later stage…)

Assumption: – zero-suppressed IT-like detector (3072 channels per Tell1 board)

– 20 tracks per primary vertex per plane (conservative)

– Each hit is a double-strip cluster encoded in 12 bits (channel address) + 3 bits for interstrip distance

Data size per vertex (one ring):

15 bits/clus x 20 clus/plane x 8 planes = 300 Bytes

Data rate:

300 Bytes x 100 Hz/bunch x 2800 bunches = 84 MB/s

Say: ~ 100 MB/s ZS cluster data ~ few hundred TB/yr

(no trigger selection, no data reduction by online processing)

+ noise!+ noise!

+ bkg!+ headers…+ bkg!+ headers…

Raw (good) event rate ~ 280 kHz


Detector cost estimate: Silicon «à la LHCb Upgrade»Detector cost estimate: Silicon «à la LHCb Upgrade»

Item kCHF Quantity Remark

Sensors ~1100 768 2x8x(3x4x4) sensors

FE electronics ~400 768 / 256 Chips / Hybrids

Readout chain ~200 4 / 384 / 384 Tell40 / GBT links / fibers

HV/LV ~200

Mech./cooling ~100

TOTAL ~2000

2 rings2 rings 8 planes8 planes

1 plane = 48 sensors•ladder = 3 sensors bonded•3 chips per ladder•1 chip = 128 channels

1 plane = 48 sensors•ladder = 3 sensors bonded•3 chips per ladder•1 chip = 128 channels

no spares taken into account=> should add 10-15% costno spares taken into account=> should add 10-15% cost

~30x30cm2~30x30cm2

Quick & dirty estimationQuick & dirty estimation


Detector cost estimate: SciFi «à la LHCb Upgrade»Detector cost estimate: SciFi «à la LHCb Upgrade»

Item kCHF Quantity Remark

Fiber ~20 47 km 2x8x(4x0.384mx5x12x128x0.25)

FE electronics ~1000 768 / 768 SiPM / Chips (+Hybrids)

Readout chain ~150 4 / 384 / 384 Tell40 / GBT links / fibers

HV/LV ~200

Mech./cooling ~100

TOTAL ~1500

2 rings2 rings 8 planes8 planes

1 plane = 4x12x128 channels(1 SiPM = 128 channels)

1 plane = 4x12x128 channels(1 SiPM = 128 channels)

no spares taken into account=> should add 10-15% costno spares taken into account=> should add 10-15% cost

~38.4x38.4cm2~38.4x38.4cm2

1 plane = 5-fiber layer, 0.25mm diameter1 plane = 5-fiber layer, 0.25mm diameter

0.25mm0.25mm

12 SiPM

NB1: cost almost doesn’t change with increasing size of detectorNB2: hard to get fiber diameter down...

Quick & dirty estimationQuick & dirty estimation


gas targetgas target


The gas targetThe gas target

Requirements:Requirements:must give sufficient beam-gas rate from the nominal gas z-range:

R(nom. range) > ~100Hz/1011 p

and sufficiently low bkg (from non-nominal z range)

R(bkg) < R(nom. range) – Also: minimize «useless» radiation damage to detector

largest possible track multiplicity per vertex– Xe better than Ne better than He ...

must be able to operate continously (when beam in machine)no effect on beam operation

– lifetime due to injected gas >100h

– contamination of nearby section to be kept low

– check SEY for gettered or cryosorbed gas species


Which gas ?Which gas ?

Larger A larger cross section and larger multiplicity per vertex

Getterable / non-getterable ? (or cryosorption ?)

getterable: e.g. CO2, N2, O2 ...

– Very local pressure bump

– Requires regenerating (changing ?) the NEG sporadically

non-getterable: e.g. Ne, Xe, …– Longer pressure bump, requires differential pumping around the target

– Some contamination of nearby cryo section surfaces ?

Charged pion multiplicities

MpA ≈ Mpp∙ (a+b∙A1/3)

a≈0.65, b≈0.3

Charged pion multiplicities

MpA ≈ Mpp∙ (a+b∙A1/3)

a≈0.65, b≈0.3

Inelastic cross section

pA ≈ pp∙ A2/3

Inelastic cross section

pA ≈ pp∙ A2/3

MG, DR, Z. Phys C65, 215-223 (1995)MG, DR, Z. Phys C65, 215-223 (1995)


Gas target (sketch!!)Gas target (sketch!!)gas injection and accumulation

pumping pumping

WF suppression

cold surface (increase density for same amount of injected gas)

detector

detector

long (small diameter) pipes act as flow restrictions

ideal pressure profile

z

p

more realistic pressure profile

this must be transparent! (thin)

beam

Reminder: contrary to VELO/LHCb, here gas injection must be continuous (with beam)... Consider impact on machine!Reminder: contrary to VELO/LHCb, here gas injection must be continuous (with beam)... Consider impact on machine!


organizationorganization


Three (parallel) activitiesThree (parallel) activities

«Off-the-trunk» device: (immediate)– install asap in LHC (LS1 or winter stop thereafter)

– proof of principle, gain experience

– only one LHC ring ?

Full-blown solution for LHC: (mid-term)– enough detectors (redundancy, acceptance)

– one device per per ring in LS2

– buy new or recycle LHCb ?

Further applications: (long-term)– SPS, PS, etc.


Next steps: tentative listNext steps: tentative list

Write down specifications of the new imaging device for LHC– detector

– gas target

Find best place in LHC ( smallest Rpipe/beam )

Organization: who works on what ?– The device is mostly a beam diagnostics instrument project led by

BE-BI, with strong support from TE-VSC, BE-ABP, LHC-OP, LHC-MP and PH

Detector•Mechanics•Sensors•FE electronics•RO board•Ctrl electronics•Ctrl s/w•Acq f/w•Reconstr s/w•Monitoring s/w•...

Detector•Mechanics•Sensors•FE electronics•RO board•Ctrl electronics•Ctrl s/w•Acq f/w•Reconstr s/w•Monitoring s/w•...

Gas Target•Vacuum sys•Gas injection•Cryo cooling•Impedance•Ctrl s/w•Monitoring s/w•...

Gas Target•Vacuum sys•Gas injection•Cryo cooling•Impedance•Ctrl s/w•Monitoring s/w•...


We need a (generic) name for such a device !!We need a (generic) name for such a device !!

BGI alas, already in use...

BSID Beam Shape Imaging Detector (Device)

VIDET Vertex Imaging DETector videt = latin for «it sees»

AVID Accelerator Vertex Imaging Detector

VID Vertex Imaging Detector (Device)

APID Accelerator Particle? Imaging Detector

VELO VErtex LOcator

BAGI Beam Analyzer with Gas Interactions

???

???

???

???

Beam

Imaging

Locator

Vertex

InteractionGas

DeviceDetector

Shape

Reconstructor

Analyzer

Accelerator


backupbackup


Option 2 for the detector arrangement (one XY plane)Option 2 for the detector arrangement (one XY plane)

plane 1 (X)plane 1 (X) plane 2 (Y)plane 2 (Y)


Option 3 for the detector arrangement (one XY plane)Option 3 for the detector arrangement (one XY plane)

Two modules types (two lengths)



Option1 for the detector arrangement (one XY plane)Option1 for the detector arrangement (one XY plane)


electronics

one sensor

number of sensors in one square is arbitrary (as long as they form a square)

number of sensors in one square is arbitrary (as long as they form a square)


LHCb electronics architectureLHCb electronics architecture

a new lhc device for the lhc: implementation aspects/options/cost estimates

Documents